This study will assess the influence of visual information on physician judgments regarding patients with potential acute ischemic heart diseases (AIHD). We will develop "cases" consisting of visual information (in the form of 30-45 second videotapes of actual patients presenting to an emergency room) and written descriptions containing only historical and laboratory information. These cases will include patients at three specified probability levels of potential AIHD; 0-10%, 20-30%, and 50-60%. The Acute Ischemic Heart Disease Predictive Instrument (AIHDPI) will be a "gold standard" for the study by determining the objective probability levels of potential AIHD in patients presenting to the emergency room. To explore the role of visual information in medical decision making, we will use an experimental design in which visual information and written descriptions, both matched and mismatched for probability ranges, will be shown to two groups of physicians. one group will read the written description and: 1) estimate the probability of AIHD, 2) place confidence intervals about their probability estimate, and 3) decide for or against admission to the coronary care unit. Then they will watch a videotape of a patient with potential AIHD and repeat the measures. A second group of physicians will do the same study except that they will see the videotapes first and written descriptions second. The significance of this research rests on the important but largely unstudied role of visual information in medical decision making. Specifically, the project derives significance from the documented overestimation of disease probabilities in patients with potential AIHD. In addition to examining the effects of visual information on physician decision making in the context of AIHD, this study will explore the effects of order of presentation of data and physician confidence in probability estimation.